To find synonyms, definitions and example sentences using WordNet

Question

I need to take an input text file with a one word. I then need to find the lemma_names, definition and examples of the synset of the word using wordnet. I have gone through the book : "Python Text Processing with NLTK 2.0 Cookbook" and also "Natural Language Processing using NLTK" to help me in this direction. Though I have understood how this can be done using the terminal, I'm not able to do the same using a text editor.

For example, if the input text has the word "flabbergasted", the output needs to be in this fashion:

flabbergasted (verb) flabbergast, boggle, bowl over - overcome with amazement ; "This boggles the mind!" (adjective) dumbfounded , dumfounded , flabbergasted , stupefied , thunderstruck , dumbstruck , dumbstricken - as if struck dumb with astonishment and surprise; "a circle of policement stood dumbfounded by her denial of having seen the accident"; "the flabbergasted aldermen were speechless"; "was thunderstruck by the news of his promotion"

The synsets, definitions and example sentences are obtained from WordNet directly!

I have the following piece of code:

from __future__ import division
import nltk
from nltk.corpus import wordnet as wn


tokenizer = nltk.data.load('tokenizers/punkt/english.pickle')
fp = open("inpsyn.txt")
data = fp.read()

#to tokenize input text into sentences

print '\n-----\n'.join(tokenizer.tokenize(data))# splits text into sentences

#to tokenize the tokenized sentences into words

tokens = nltk.wordpunct_tokenize(data)
text = nltk.Text(tokens)
words = [w.lower() for w in text]  
print words     #to print the tokens

for a in words:
    print a

syns = wn.synsets(a)
print "synsets:", syns

for s in syns:
    for l in s.lemmas:
        print l.name
    print s.definition
    print s.examples

I get the following output:

flabbergasted

['flabbergasted']
flabbergasted
synsets: [Synset('flabbergast.v.01'), Synset('dumbfounded.s.01')]
flabbergast
boggle
bowl_over
overcome with amazement
['This boggles the mind!']
dumbfounded
dumfounded
flabbergasted
stupefied
thunderstruck
dumbstruck
dumbstricken
as if struck dumb with astonishment and surprise
['a circle of policement stood dumbfounded by her denial of having seen the accident', 'the flabbergasted aldermen were speechless', 'was thunderstruck by the news of his promotion']

Is there a way to retrieve the part of speech along with the group of lemma names?

Answer 1

def synset(word):
    wn.synsets(word)

doesn't return anything so by default you get None

you should write

def synset(word):
    return wn.synsets(word)

Extracting lemma names:

from nltk.corpus import wordnet
syns = wordnet.synsets('car')
syns[0].lemmas[0].name
>>> 'car'
[s.lemmas[0].name for s in syns]
>>> ['car', 'car', 'car', 'car', 'cable_car']


[l.name for s in syns for l in s.lemmas]
>>>['car', 'auto', 'automobile', 'machine', 'motorcar', 'car', 'railcar', 'railway_car', 'railroad_car', 'car', 'gondola', 'car', 'elevator_car', 'cable_car', 'car']

Answer 2

Here I have created a module which can easily be used(imported), and with a string being passed to it, will return all the lemma words of the string.

Module:

#!/usr/bin/python2.7
''' pass a string to this funciton ( eg 'car') and it will give you a list of
words which is related to cat, called lemma of CAT. '''
from nltk.corpus import wordnet as wn
import sys
#print all the synset element of an element
def lemmalist(str):
    syn_set = []
    for synset in wn.synsets(str):
        for item in synset.lemma_names:
            syn_set.append(item)
    return syn_set

Usage:

Note: module name is lemma.py hence "from lemma import lemmalist"

>>> from lemma import lemmalist
>>> lemmalist('car')
['car', 'auto', 'automobile', 'machine', 'motorcar', 'car', 'railcar', 'railway_car', 'railroad_car', 'car', 'gondola', 'car', 'elevator_car', 'cable_car', 'car']

Cheers!

Answer 3

synonyms = []
for syn in wordnet.synsets("car"):
    for l in syn.lemmas():
        synonyms.append(l.name())
print synonyms

Answer 4

In NLTK 3.0, lemma_names has been changed from attribute to method. So if you get an error saying:

TypeError: 'method' object is not iterable

You can fix it using:

>>> from nltk.corpus import wordnet as wn
>>> [item for sysnet in wn.synsets('car') for item in sysnet.lemma_names()]

This will output:

>>> [
       'car', 'auto', 'automobile', 'machine', 'motorcar', 'car', 
       'railcar', 'railway_car', 'railroad_car', 'car', 'gondola', 
       'car', 'elevator_car', 'cable_car', 'car'
    ]